root / doc / admin.sgml @ 7c18ef8e
History | View | Annotate | Download (14 kB)
1 | 6bf0c9bb | Guido Trotter | <!DOCTYPE article PUBLIC "-//OASIS//DTD DocBook V4.2//EN" [ |
---|---|---|---|
2 | 6bf0c9bb | Guido Trotter | ]> |
3 | 6bf0c9bb | Guido Trotter | <article class="specification"> |
4 | 6bf0c9bb | Guido Trotter | <articleinfo> |
5 | 6bf0c9bb | Guido Trotter | <title>Ganeti administrator's guide</title> |
6 | 6bf0c9bb | Guido Trotter | </articleinfo> |
7 | 6bf0c9bb | Guido Trotter | <para>Documents Ganeti version 1.2</para> |
8 | 6bf0c9bb | Guido Trotter | <sect1> |
9 | 6bf0c9bb | Guido Trotter | <title>Introduction</title> |
10 | 6bf0c9bb | Guido Trotter | |
11 | 6bf0c9bb | Guido Trotter | <para>Ganeti is a virtualization cluster management software. You are |
12 | 6bf0c9bb | Guido Trotter | expected to be a system administrator familiar with your Linux distribution |
13 | 6bf0c9bb | Guido Trotter | and the Xen virtualization environment before using it. |
14 | 6bf0c9bb | Guido Trotter | </para> |
15 | 6bf0c9bb | Guido Trotter | |
16 | 6bf0c9bb | Guido Trotter | <para>The various components of Ganeti all have man pages and interactive |
17 | 6bf0c9bb | Guido Trotter | help. This manual though will help you getting familiar with the system by |
18 | 6bf0c9bb | Guido Trotter | explaining the most common operations, grouped by related use. |
19 | 6bf0c9bb | Guido Trotter | </para> |
20 | 6bf0c9bb | Guido Trotter | |
21 | 6bf0c9bb | Guido Trotter | <para>After a terminology glossary and a section on the prerequisites |
22 | 6bf0c9bb | Guido Trotter | needed to use this manual, the rest of this document is divided in three |
23 | 6bf0c9bb | Guido Trotter | main sections, which group different features of Ganeti: |
24 | 6bf0c9bb | Guido Trotter | <itemizedlist> |
25 | 6bf0c9bb | Guido Trotter | <listitem> |
26 | 6bf0c9bb | Guido Trotter | <simpara>Instance Management</simpara> |
27 | 6bf0c9bb | Guido Trotter | </listitem> |
28 | 6bf0c9bb | Guido Trotter | <listitem> |
29 | 6bf0c9bb | Guido Trotter | <simpara>High Availability Features</simpara> |
30 | 6bf0c9bb | Guido Trotter | </listitem> |
31 | 6bf0c9bb | Guido Trotter | <listitem> |
32 | 6bf0c9bb | Guido Trotter | <simpara>Debugging Features</simpara> |
33 | 6bf0c9bb | Guido Trotter | </listitem> |
34 | 6bf0c9bb | Guido Trotter | </itemizedlist> |
35 | 6bf0c9bb | Guido Trotter | </para> |
36 | 6bf0c9bb | Guido Trotter | |
37 | 6bf0c9bb | Guido Trotter | <sect2> |
38 | ec377077 | Iustin Pop | |
39 | ec377077 | Iustin Pop | <title>Ganeti terminology</title> |
40 | 6bf0c9bb | Guido Trotter | |
41 | 6bf0c9bb | Guido Trotter | <para>This section provides a small introduction to Ganeti terminology, |
42 | 6bf0c9bb | Guido Trotter | which might be useful to read the rest of the document. |
43 | 6bf0c9bb | Guido Trotter | |
44 | ec377077 | Iustin Pop | <glosslist> |
45 | ec377077 | Iustin Pop | <glossentry> |
46 | ec377077 | Iustin Pop | <glossterm>Cluster</glossterm> |
47 | ec377077 | Iustin Pop | <glossdef> |
48 | ec377077 | Iustin Pop | <simpara> |
49 | ec377077 | Iustin Pop | A set of machines (nodes) that cooperate to offer a |
50 | ec377077 | Iustin Pop | coherent highly available virtualization service. |
51 | ec377077 | Iustin Pop | </simpara> |
52 | ec377077 | Iustin Pop | </glossdef> |
53 | ec377077 | Iustin Pop | </glossentry> |
54 | ec377077 | Iustin Pop | <glossentry> |
55 | ec377077 | Iustin Pop | <glossterm>Node</glossterm> |
56 | ec377077 | Iustin Pop | <glossdef> |
57 | ec377077 | Iustin Pop | <simpara> |
58 | ec377077 | Iustin Pop | A physical machine which is member of a cluster. |
59 | ec377077 | Iustin Pop | Nodes are the basic cluster infrastructure, and are |
60 | ec377077 | Iustin Pop | not fault tolerant. |
61 | ec377077 | Iustin Pop | </simpara> |
62 | ec377077 | Iustin Pop | </glossdef> |
63 | ec377077 | Iustin Pop | </glossentry> |
64 | ec377077 | Iustin Pop | <glossentry> |
65 | ec377077 | Iustin Pop | <glossterm>Master node</glossterm> |
66 | ec377077 | Iustin Pop | <glossdef> |
67 | ec377077 | Iustin Pop | <simpara> |
68 | ec377077 | Iustin Pop | The node which controls the Cluster, from which all |
69 | ec377077 | Iustin Pop | Ganeti commands must be given. |
70 | ec377077 | Iustin Pop | </simpara> |
71 | ec377077 | Iustin Pop | </glossdef> |
72 | ec377077 | Iustin Pop | </glossentry> |
73 | ec377077 | Iustin Pop | <glossentry> |
74 | ec377077 | Iustin Pop | <glossterm>Instance</glossterm> |
75 | ec377077 | Iustin Pop | <glossdef> |
76 | ec377077 | Iustin Pop | <simpara> |
77 | ec377077 | Iustin Pop | A virtual machine which runs on a cluster. It can be a |
78 | ec377077 | Iustin Pop | fault tolerant highly available entity. |
79 | ec377077 | Iustin Pop | </simpara> |
80 | ec377077 | Iustin Pop | </glossdef> |
81 | ec377077 | Iustin Pop | </glossentry> |
82 | ec377077 | Iustin Pop | <glossentry> |
83 | ec377077 | Iustin Pop | <glossterm>Pool</glossterm> |
84 | ec377077 | Iustin Pop | <glossdef> |
85 | ec377077 | Iustin Pop | <simpara> |
86 | ec377077 | Iustin Pop | A pool is a set of clusters sharing the same network. |
87 | ec377077 | Iustin Pop | </simpara> |
88 | ec377077 | Iustin Pop | </glossdef> |
89 | ec377077 | Iustin Pop | </glossentry> |
90 | ec377077 | Iustin Pop | <glossentry> |
91 | ec377077 | Iustin Pop | <glossterm>Meta-Cluster</glossterm> |
92 | ec377077 | Iustin Pop | <glossdef> |
93 | ec377077 | Iustin Pop | <simpara> |
94 | ec377077 | Iustin Pop | Anything that concerns more than one cluster. |
95 | ec377077 | Iustin Pop | </simpara> |
96 | ec377077 | Iustin Pop | </glossdef> |
97 | ec377077 | Iustin Pop | </glossentry> |
98 | ec377077 | Iustin Pop | </glosslist> |
99 | 6bf0c9bb | Guido Trotter | |
100 | 6bf0c9bb | Guido Trotter | </para> |
101 | 6bf0c9bb | Guido Trotter | </sect2> |
102 | 6bf0c9bb | Guido Trotter | |
103 | 6bf0c9bb | Guido Trotter | <sect2> |
104 | 6bf0c9bb | Guido Trotter | <title>Prerequisites</title> |
105 | 6bf0c9bb | Guido Trotter | |
106 | ec377077 | Iustin Pop | <para> |
107 | ec377077 | Iustin Pop | You need to have your Ganeti cluster installed and configured |
108 | ec377077 | Iustin Pop | before you try any of the commands in this document. Please |
109 | ec377077 | Iustin Pop | follow the <emphasis>Ganeti installation tutorial</emphasis> |
110 | ec377077 | Iustin Pop | for instructions on how to do that. |
111 | 6bf0c9bb | Guido Trotter | </para> |
112 | 6bf0c9bb | Guido Trotter | </sect2> |
113 | 6bf0c9bb | Guido Trotter | |
114 | 6bf0c9bb | Guido Trotter | </sect1> |
115 | 6bf0c9bb | Guido Trotter | |
116 | 6bf0c9bb | Guido Trotter | <sect1> |
117 | 6bf0c9bb | Guido Trotter | <title>Managing Instances</title> |
118 | 6bf0c9bb | Guido Trotter | |
119 | 6bf0c9bb | Guido Trotter | <sect2> |
120 | 6bf0c9bb | Guido Trotter | <title>Adding/Removing an instance</title> |
121 | 6bf0c9bb | Guido Trotter | |
122 | ec377077 | Iustin Pop | <para> |
123 | ec377077 | Iustin Pop | Adding a new virtual instance to your Ganeti cluster is really |
124 | ec377077 | Iustin Pop | easy. The command is: |
125 | ec377077 | Iustin Pop | |
126 | ec377077 | Iustin Pop | <synopsis>gnt-instance add -n <replaceable>TARGET_NODE</replaceable> -o <replaceable>OS_TYPE</replaceable> -t <replaceable>DISK_TEMPLATE</replaceable> <replaceable>INSTANCE_NAME</replaceable></synopsis> |
127 | ec377077 | Iustin Pop | |
128 | ec377077 | Iustin Pop | The instance name must be resolvable (e.g. exist in DNS) and |
129 | ec377077 | Iustin Pop | of course map to an address in the same subnet as the cluster |
130 | ec377077 | Iustin Pop | itself. Options you can give to this command include: |
131 | ec377077 | Iustin Pop | |
132 | 6bf0c9bb | Guido Trotter | <itemizedlist> |
133 | 6bf0c9bb | Guido Trotter | <listitem> |
134 | ec377077 | Iustin Pop | <simpara>The disk size (<option>-s</option>)</simpara> |
135 | 6bf0c9bb | Guido Trotter | </listitem> |
136 | 6bf0c9bb | Guido Trotter | <listitem> |
137 | ec377077 | Iustin Pop | <simpara>The swap size (<option>--swap-size</option>)</simpara> |
138 | 6bf0c9bb | Guido Trotter | </listitem> |
139 | 6bf0c9bb | Guido Trotter | <listitem> |
140 | ec377077 | Iustin Pop | <simpara>The memory size (<option>-m</option>)</simpara> |
141 | 6bf0c9bb | Guido Trotter | </listitem> |
142 | 6bf0c9bb | Guido Trotter | <listitem> |
143 | ec377077 | Iustin Pop | <simpara>The number of virtual CPUs (<option>-p</option>)</simpara> |
144 | 6bf0c9bb | Guido Trotter | </listitem> |
145 | 6bf0c9bb | Guido Trotter | <listitem> |
146 | ec377077 | Iustin Pop | <simpara>The instance ip address (<option>-i</option>) (use |
147 | ec377077 | Iustin Pop | the value <literal>auto</literal> to make Ganeti record the |
148 | ec377077 | Iustin Pop | address from dns)</simpara> |
149 | 6bf0c9bb | Guido Trotter | </listitem> |
150 | 6bf0c9bb | Guido Trotter | <listitem> |
151 | ec377077 | Iustin Pop | <simpara>The bridge to connect the instance to |
152 | ec377077 | Iustin Pop | (<option>-b</option>), if you don't want to use the default |
153 | ec377077 | Iustin Pop | one</simpara> |
154 | 6bf0c9bb | Guido Trotter | </listitem> |
155 | 6bf0c9bb | Guido Trotter | </itemizedlist> |
156 | 53b78ba4 | Guido Trotter | </para> |
157 | 53b78ba4 | Guido Trotter | |
158 | ec377077 | Iustin Pop | <para>There are four types of disk template you can choose from:</para> |
159 | 53b78ba4 | Guido Trotter | |
160 | 53b78ba4 | Guido Trotter | <variablelist> |
161 | 53b78ba4 | Guido Trotter | <varlistentry> |
162 | 53b78ba4 | Guido Trotter | <term>diskless</term> |
163 | 53b78ba4 | Guido Trotter | <listitem><para>The instance has no disks. Only used for special |
164 | 53b78ba4 | Guido Trotter | purpouse operating systems or for testing.</para></listitem> |
165 | 53b78ba4 | Guido Trotter | </varlistentry> |
166 | 53b78ba4 | Guido Trotter | |
167 | 53b78ba4 | Guido Trotter | <varlistentry> |
168 | 53b78ba4 | Guido Trotter | <term>plain</term> |
169 | 53b78ba4 | Guido Trotter | <listitem><para>The instance will use LVM devices as backend for its |
170 | 53b78ba4 | Guido Trotter | disks. No redundancy is provided.</para></listitem> |
171 | 53b78ba4 | Guido Trotter | </varlistentry> |
172 | 53b78ba4 | Guido Trotter | |
173 | 53b78ba4 | Guido Trotter | <varlistentry> |
174 | 53b78ba4 | Guido Trotter | <term>local_raid1</term> |
175 | 53b78ba4 | Guido Trotter | <listitem><para>A local mirror is set between LVM devices to back the |
176 | 53b78ba4 | Guido Trotter | instance. This provides some redundancy for the instance's |
177 | 53b78ba4 | Guido Trotter | data.</para></listitem> |
178 | 53b78ba4 | Guido Trotter | </varlistentry> |
179 | 53b78ba4 | Guido Trotter | |
180 | 53b78ba4 | Guido Trotter | <varlistentry> |
181 | 53b78ba4 | Guido Trotter | <term>remote_raid1</term> |
182 | ec377077 | Iustin Pop | <listitem> |
183 | ec377077 | Iustin Pop | <simpara><emphasis role="strong">Note:</emphasis> This is |
184 | ec377077 | Iustin Pop | only valid for multi-node clusters.</simpara> |
185 | ec377077 | Iustin Pop | <simpara> |
186 | ec377077 | Iustin Pop | A mirror is set between the local node and a remote |
187 | ec377077 | Iustin Pop | one, which must be specified with the --secondary-node |
188 | ec377077 | Iustin Pop | option. Use this option to obtain a highly available |
189 | ec377077 | Iustin Pop | instance that can be failed over to a remote node |
190 | ec377077 | Iustin Pop | should the primary one fail. |
191 | ec377077 | Iustin Pop | </simpara> |
192 | ec377077 | Iustin Pop | </listitem> |
193 | 53b78ba4 | Guido Trotter | </varlistentry> |
194 | 53b78ba4 | Guido Trotter | |
195 | 53b78ba4 | Guido Trotter | </variablelist> |
196 | 53b78ba4 | Guido Trotter | |
197 | ec377077 | Iustin Pop | <para> |
198 | ec377077 | Iustin Pop | For example if you want to create an highly available instance |
199 | ec377077 | Iustin Pop | use the remote_raid1 disk template: |
200 | ec377077 | Iustin Pop | <synopsis>gnt-instance add -n <replaceable>TARGET_NODE</replaceable> -o <replaceable>OS_TYPE</replaceable> -t remote_raid1 \ |
201 | ec377077 | Iustin Pop | --secondary-node=<replaceable>SECONDARY_NODE</replaceable> <replaceable>INSTANCE_NAME</replaceable></synopsis> |
202 | ec377077 | Iustin Pop | |
203 | ec377077 | Iustin Pop | <para> |
204 | ec377077 | Iustin Pop | To know which operating systems your cluster supports you can use: |
205 | ec377077 | Iustin Pop | |
206 | ec377077 | Iustin Pop | <synopsis>gnt-os list</synopsis> |
207 | ec377077 | Iustin Pop | |
208 | 6bf0c9bb | Guido Trotter | </para> |
209 | 6bf0c9bb | Guido Trotter | |
210 | 6bf0c9bb | Guido Trotter | <para> |
211 | ec377077 | Iustin Pop | Removing an instance is even easier than creating one. This |
212 | ec377077 | Iustin Pop | operation is non-reversible and destroys all the contents of |
213 | ec377077 | Iustin Pop | your instance. Use with care: |
214 | ec377077 | Iustin Pop | |
215 | ec377077 | Iustin Pop | <synopsis>gnt-instance remove <replaceable>INSTANCE_NAME</replaceable></synopsis> |
216 | ec377077 | Iustin Pop | |
217 | 6bf0c9bb | Guido Trotter | </para> |
218 | 6bf0c9bb | Guido Trotter | </sect2> |
219 | 6bf0c9bb | Guido Trotter | |
220 | 6bf0c9bb | Guido Trotter | <sect2> |
221 | 6bf0c9bb | Guido Trotter | <title>Starting/Stopping an instance</title> |
222 | 6bf0c9bb | Guido Trotter | |
223 | ec377077 | Iustin Pop | <para> |
224 | ec377077 | Iustin Pop | Instances are automatically started at instance creation |
225 | ec377077 | Iustin Pop | time. To manually start one which is currently stopped you can |
226 | ec377077 | Iustin Pop | run: |
227 | ec377077 | Iustin Pop | |
228 | ec377077 | Iustin Pop | <synopsis>gnt-instance startup <replaceable>INSTANCE_NAME</replaceable></synopsis> |
229 | ec377077 | Iustin Pop | |
230 | ec377077 | Iustin Pop | While the command to stop one is: |
231 | ec377077 | Iustin Pop | |
232 | ec377077 | Iustin Pop | <synopsis>gnt-instance shutdown <replaceable>INSTANCE_NAME</replaceable></synopsis> |
233 | ec377077 | Iustin Pop | |
234 | ec377077 | Iustin Pop | The command to see all the instances configured and their |
235 | ec377077 | Iustin Pop | status is: |
236 | ec377077 | Iustin Pop | |
237 | ec377077 | Iustin Pop | <synopsis>gnt-instance list</synopsis> |
238 | ec377077 | Iustin Pop | |
239 | 6bf0c9bb | Guido Trotter | </para> |
240 | 53b78ba4 | Guido Trotter | |
241 | ec377077 | Iustin Pop | <para> |
242 | ec377077 | Iustin Pop | Do not use the xen commands to stop instances. If you run for |
243 | ec377077 | Iustin Pop | example xm shutdown or xm destroy on an instance Ganeti will |
244 | ec377077 | Iustin Pop | automatically restart it (via the |
245 | ec377077 | Iustin Pop | <citerefentry><refentrytitle>ganeti-watcher</refentrytitle> |
246 | ec377077 | Iustin Pop | <manvolnum>8</manvolnum></citerefentry>) |
247 | 53b78ba4 | Guido Trotter | </para> |
248 | 53b78ba4 | Guido Trotter | |
249 | 6bf0c9bb | Guido Trotter | </sect2> |
250 | 6bf0c9bb | Guido Trotter | |
251 | 6bf0c9bb | Guido Trotter | <sect2> |
252 | 6bf0c9bb | Guido Trotter | <title>Exporting/Importing an instance</title> |
253 | 6bf0c9bb | Guido Trotter | |
254 | ec377077 | Iustin Pop | <para> |
255 | ec377077 | Iustin Pop | You can create a snapshot of an instance disk and Ganeti |
256 | ec377077 | Iustin Pop | configuration, which then you can backup, or import into |
257 | ec377077 | Iustin Pop | another cluster. The way to export an instance is: |
258 | ec377077 | Iustin Pop | |
259 | ec377077 | Iustin Pop | <synopsis>gnt-backup export -n <replaceable>TARGET_NODE</replaceable> <replaceable>INSTANCE_NAME</replaceable></synopsis> |
260 | ec377077 | Iustin Pop | |
261 | ec377077 | Iustin Pop | The target node can be any node in the cluster with enough |
262 | ec377077 | Iustin Pop | space under <filename class="directory">/srv/ganeti</filename> |
263 | ec377077 | Iustin Pop | to hold the instance image. Use the |
264 | ec377077 | Iustin Pop | <option>--noshutdown</option> option to snapshot an instance |
265 | ec377077 | Iustin Pop | without rebooting it. Any previous snapshot of the same |
266 | ec377077 | Iustin Pop | instance existing cluster-wide under <filename |
267 | ec377077 | Iustin Pop | class="directory">/srv/ganeti</filename> will be removed by |
268 | ec377077 | Iustin Pop | this operation: if you want to keep them move them out of the |
269 | ec377077 | Iustin Pop | Ganeti exports directory. |
270 | 6bf0c9bb | Guido Trotter | </para> |
271 | 6bf0c9bb | Guido Trotter | |
272 | ec377077 | Iustin Pop | <para> |
273 | ec377077 | Iustin Pop | Importing an instance is similar to creating a new one. The |
274 | ec377077 | Iustin Pop | command is: |
275 | ec377077 | Iustin Pop | |
276 | ec377077 | Iustin Pop | <synopsis>gnt-backup import -n <replaceable>TARGET_NODE</replaceable> -t <replaceable>DISK_TEMPLATE</replaceable> --src-node=<replaceable>NODE</replaceable> --src-dir=DIR INSTANCE_NAME</synopsis> |
277 | ec377077 | Iustin Pop | |
278 | ec377077 | Iustin Pop | Most of the options available for the command |
279 | ec377077 | Iustin Pop | <emphasis>gnt-instance add</emphasis> are supported here too. |
280 | ec377077 | Iustin Pop | |
281 | 6bf0c9bb | Guido Trotter | </para> |
282 | 6bf0c9bb | Guido Trotter | </sect2> |
283 | 6bf0c9bb | Guido Trotter | |
284 | 6bf0c9bb | Guido Trotter | </sect1> |
285 | 6bf0c9bb | Guido Trotter | |
286 | 6bf0c9bb | Guido Trotter | |
287 | 6bf0c9bb | Guido Trotter | <sect1> |
288 | 6bf0c9bb | Guido Trotter | <title>High availability features</title> |
289 | 6bf0c9bb | Guido Trotter | |
290 | ec377077 | Iustin Pop | <note> |
291 | ec377077 | Iustin Pop | <simpara>This section only applies to multi-node clusters.</simpara> |
292 | ec377077 | Iustin Pop | </note> |
293 | ec377077 | Iustin Pop | |
294 | 6bf0c9bb | Guido Trotter | <sect2> |
295 | 6bf0c9bb | Guido Trotter | <title>Failing over an instance</title> |
296 | 6bf0c9bb | Guido Trotter | |
297 | ec377077 | Iustin Pop | <para> |
298 | ec377077 | Iustin Pop | If an instance is built in highly available mode you can at |
299 | ec377077 | Iustin Pop | any time fail it over to its secondary node, even if the |
300 | ec377077 | Iustin Pop | primary has somehow failed and it's not up anymore. Doing it |
301 | ec377077 | Iustin Pop | is really easy, on the master node you can just run: |
302 | ec377077 | Iustin Pop | |
303 | ec377077 | Iustin Pop | <synopsis>gnt-instance failover <replaceable>INSTANCE_NAME</replaceable></synopsis> |
304 | ec377077 | Iustin Pop | |
305 | ec377077 | Iustin Pop | That's it. After the command completes the secondary node is |
306 | ec377077 | Iustin Pop | now the primary, and vice versa. |
307 | 6bf0c9bb | Guido Trotter | </para> |
308 | 6bf0c9bb | Guido Trotter | </sect2> |
309 | ec377077 | Iustin Pop | |
310 | 6bf0c9bb | Guido Trotter | <sect2> |
311 | 6bf0c9bb | Guido Trotter | <title>Replacing an instance disks</title> |
312 | 6bf0c9bb | Guido Trotter | |
313 | ec377077 | Iustin Pop | <para> |
314 | ec377077 | Iustin Pop | So what if instead the secondary node for an instance has |
315 | ec377077 | Iustin Pop | failed, or you plan to remove a node from your cluster, and |
316 | ec377077 | Iustin Pop | you failed over all its instances, but it's still secondary |
317 | ec377077 | Iustin Pop | for some? The solution here is to replace the instance disks, |
318 | ec377077 | Iustin Pop | changing the secondary node: |
319 | ec377077 | Iustin Pop | |
320 | ec377077 | Iustin Pop | <synopsis>gnt-instance replace-disks -n <replaceable>NEW_SECONDARY</replaceable> <replaceable>INSTANCE_NAME</replaceable></synopsis> |
321 | ec377077 | Iustin Pop | |
322 | ec377077 | Iustin Pop | This process is a bit longer, but involves no instance |
323 | ec377077 | Iustin Pop | downtime, and at the end of it the instance has changed its |
324 | ec377077 | Iustin Pop | secondary node, to which it can if necessary be failed over. |
325 | 6bf0c9bb | Guido Trotter | </para> |
326 | 6bf0c9bb | Guido Trotter | </sect2> |
327 | 6bf0c9bb | Guido Trotter | <sect2> |
328 | 6bf0c9bb | Guido Trotter | <title>Failing over the master node</title> |
329 | 6bf0c9bb | Guido Trotter | |
330 | ec377077 | Iustin Pop | <para> |
331 | ec377077 | Iustin Pop | This is all good as long as the Ganeti Master Node is |
332 | ec377077 | Iustin Pop | up. Should it go down, or should you wish to decommission it, |
333 | ec377077 | Iustin Pop | just run on any other node the command: |
334 | ec377077 | Iustin Pop | |
335 | ec377077 | Iustin Pop | <synopsis>gnt-cluster masterfailover</synopsis> |
336 | ec377077 | Iustin Pop | |
337 | ec377077 | Iustin Pop | and the node you ran it on is now the new master. |
338 | 6bf0c9bb | Guido Trotter | </para> |
339 | 6bf0c9bb | Guido Trotter | </sect2> |
340 | 6bf0c9bb | Guido Trotter | <sect2> |
341 | 6bf0c9bb | Guido Trotter | <title>Adding/Removing nodes</title> |
342 | 6bf0c9bb | Guido Trotter | |
343 | ec377077 | Iustin Pop | <para> |
344 | ec377077 | Iustin Pop | And of course, now that you know how to move instances around, |
345 | ec377077 | Iustin Pop | it's easy to free up a node, and then you can remove it from |
346 | ec377077 | Iustin Pop | the cluster: |
347 | ec377077 | Iustin Pop | |
348 | ec377077 | Iustin Pop | <synopsis> |
349 | ec377077 | Iustin Pop | gnt-node remove <replaceable>NODE_NAME</replaceable> |
350 | ec377077 | Iustin Pop | </synopsis> |
351 | ec377077 | Iustin Pop | |
352 | ec377077 | Iustin Pop | and maybe add a new one: |
353 | ec377077 | Iustin Pop | |
354 | ec377077 | Iustin Pop | <synopsis> |
355 | ec377077 | Iustin Pop | gnt-node add <optional><option>--secondary-ip=<replaceable>ADDRESS</replaceable></option></optional> <replaceable>NODE_NAME</replaceable> |
356 | ec377077 | Iustin Pop | |
357 | ec377077 | Iustin Pop | </synopsis> |
358 | 6bf0c9bb | Guido Trotter | </para> |
359 | 6bf0c9bb | Guido Trotter | </sect2> |
360 | 6bf0c9bb | Guido Trotter | </sect1> |
361 | 6bf0c9bb | Guido Trotter | |
362 | 6bf0c9bb | Guido Trotter | <sect1> |
363 | 6bf0c9bb | Guido Trotter | <title>Debugging Features</title> |
364 | 6bf0c9bb | Guido Trotter | |
365 | ec377077 | Iustin Pop | <para> |
366 | ec377077 | Iustin Pop | At some point you might need to do some debugging operations on |
367 | ec377077 | Iustin Pop | your cluster or on your instances. This section will help you |
368 | ec377077 | Iustin Pop | with the most used debugging functionalities. |
369 | 6bf0c9bb | Guido Trotter | </para> |
370 | 6bf0c9bb | Guido Trotter | |
371 | 6bf0c9bb | Guido Trotter | <sect2> |
372 | 6bf0c9bb | Guido Trotter | <title>Accessing an instance's disks</title> |
373 | 6bf0c9bb | Guido Trotter | |
374 | ec377077 | Iustin Pop | <para> |
375 | ec377077 | Iustin Pop | From an instance's primary node you have access to its |
376 | ec377077 | Iustin Pop | disks. Never ever mount the underlying logical volume manually |
377 | ec377077 | Iustin Pop | on a fault tolerant instance, or you risk breaking |
378 | ec377077 | Iustin Pop | replication. The correct way to access them is to run the |
379 | ec377077 | Iustin Pop | command: |
380 | ec377077 | Iustin Pop | |
381 | ec377077 | Iustin Pop | <synopsis> gnt-instance activate-disks <replaceable>INSTANCE_NAME</replaceable></synopsis> |
382 | ec377077 | Iustin Pop | |
383 | ec377077 | Iustin Pop | And then access the device that gets created. After you've |
384 | ec377077 | Iustin Pop | finished you can deactivate them with the deactivate-disks |
385 | ec377077 | Iustin Pop | command, which works in the same way. |
386 | 6bf0c9bb | Guido Trotter | </para> |
387 | 6bf0c9bb | Guido Trotter | </sect2> |
388 | 6bf0c9bb | Guido Trotter | |
389 | 6bf0c9bb | Guido Trotter | <sect2> |
390 | 6bf0c9bb | Guido Trotter | <title>Accessing an instance's console</title> |
391 | 6bf0c9bb | Guido Trotter | |
392 | ec377077 | Iustin Pop | <para> |
393 | ec377077 | Iustin Pop | The command to access a running instance's console is: |
394 | ec377077 | Iustin Pop | |
395 | ec377077 | Iustin Pop | <synopsis>gnt-instance console <replaceable>INSTANCE_NAME</replaceable></synopsis> |
396 | ec377077 | Iustin Pop | |
397 | ec377077 | Iustin Pop | Use the console normally and then type |
398 | ec377077 | Iustin Pop | <userinput>^]</userinput> when done, to exit. |
399 | 6bf0c9bb | Guido Trotter | </para> |
400 | 6bf0c9bb | Guido Trotter | </sect2> |
401 | 6bf0c9bb | Guido Trotter | |
402 | 6bf0c9bb | Guido Trotter | <sect2> |
403 | 6bf0c9bb | Guido Trotter | <title>Instance Operating System Debugging</title> |
404 | 6bf0c9bb | Guido Trotter | |
405 | ec377077 | Iustin Pop | <para> |
406 | ec377077 | Iustin Pop | Should you have any problems with operating systems support |
407 | ec377077 | Iustin Pop | the command to ran to see a complete status for all your nodes |
408 | ec377077 | Iustin Pop | is: |
409 | ec377077 | Iustin Pop | |
410 | ec377077 | Iustin Pop | <synopsis>gnt-os diagnose</synopsis> |
411 | ec377077 | Iustin Pop | |
412 | 6bf0c9bb | Guido Trotter | </para> |
413 | 6bf0c9bb | Guido Trotter | |
414 | 6bf0c9bb | Guido Trotter | </sect2> |
415 | 6bf0c9bb | Guido Trotter | |
416 | 6bf0c9bb | Guido Trotter | <sect2> |
417 | 6bf0c9bb | Guido Trotter | <title>Cluster-wide debugging</title> |
418 | 6bf0c9bb | Guido Trotter | |
419 | ec377077 | Iustin Pop | <para> |
420 | ec377077 | Iustin Pop | The gnt-cluster command offers several options to run tests or |
421 | ec377077 | Iustin Pop | execute cluster-wide operations. For example: |
422 | ec377077 | Iustin Pop | |
423 | ec377077 | Iustin Pop | <screen> |
424 | 6bf0c9bb | Guido Trotter | gnt-cluster command |
425 | 6bf0c9bb | Guido Trotter | gnt-cluster copyfile |
426 | 6bf0c9bb | Guido Trotter | gnt-cluster verify |
427 | 6bf0c9bb | Guido Trotter | gnt-cluster getmaster |
428 | 6bf0c9bb | Guido Trotter | gnt-cluster version |
429 | ec377077 | Iustin Pop | </screen> |
430 | ec377077 | Iustin Pop | |
431 | ec377077 | Iustin Pop | See the man page <citerefentry> |
432 | ec377077 | Iustin Pop | <refentrytitle>gnt-cluster</refentrytitle> |
433 | ec377077 | Iustin Pop | <manvolnum>8</manvolnum> </citerefentry> to know more about |
434 | ec377077 | Iustin Pop | their usage. |
435 | 6bf0c9bb | Guido Trotter | </para> |
436 | 6bf0c9bb | Guido Trotter | </sect2> |
437 | 6bf0c9bb | Guido Trotter | |
438 | 6bf0c9bb | Guido Trotter | </sect1> |
439 | 6bf0c9bb | Guido Trotter | |
440 | 6bf0c9bb | Guido Trotter | </article> |